Eric Xihui Lin

[, [[ and $

Wed, 02 Sep 2015 00:00:00 +0000

Perhaps [, [[ and $ are the most frequently used functions/operators in R for everyone. But how much do you know about them?

Behaviors for vector, matrix and data.frame

for vector

I mean, atomic vector like c(5, 1, 10) or recursive vector like list.

> x <- c(A = 'T', C = 'G', G = 'C', T = 'A')
> x
  A   C   G   T 
"T" "G" "C" "A" 

> lx <- list(AA = 1:4, BB = 3:5)
> lx
$AA
[1] 1 2 3 4
$BB
[1] 3 4 5

[ keeps attributes (names, class, etc), while [[ does not.

> x[1] # or x['A'], name kept
  A 
"T" 
> x[[1]]  # or x[['A']], name dropped
[1] "T"

> lx[1]  # or lx['AA']. still a list. Class kept
$AA
[1] 1 2 3 4

> lx[[1]]  # or lx[['AA']].  NOT a list
[1] 1 2 3 4

[ takes a vector of indices, while [[ cannot (of course as from the above property)

> x[c('A', 'T')]
  A   T 
"T" "A" 
> x[[c('A', 'T')]]
Error in x[[c("A", "T")]] : attempt to select more than one element

[ accepts negative integer to extract "all-but-these" element, while [[ does not accept negative integer.
$ can be used in recursive type like list, data frame and environment, but not atomic type like c(1,2,3). Conceptually list$i is equivalent to list[['i', exact = FALSE]], where i must be literature names (string).

> lx[['A']]  # lx has elements of AA, BB, but not A.
NULL
> lx[['A', exact = FALSE]]
[1] 1 2 3 4
> lx$A
[1] 1 2 3 4
# use `` to wrap over the name if it contains an invalid name character(e.g., space)

So obviously, you will choose $ over [[ when you want partial match or simpler syntax. But you may want to use [[ over $ if you want exact match, or if i is an expression or variable.

for matrix

First, matrix is an atomic object, but is not of a vector type.

$ cannot be used. Both [ and [[, with one argument like matrix[1] and matrix[[1]] is valid and is exactly the same, with both does not keep attributes.
matrix[vectorOfIndices] behaviours identical to as.vector(matrix)[vectorOfIndices], except when vectorOfIndices is missing (which does nothing but return the object itself)
matrix[i, j]: arguments are matched by position, not keyword, i.e., matrix[j = 2, i = 1] returns matrix[2, 1] NOT matrix[1,2].
matrix[1:3, 1] or matrix[1, 1:3] returns a vector, NOT matrix. To request a matrix, use matrix[1:3, 1, drop = FALSE], etc.

for data.frame

Data frame is an interest object, which is kind of a mix of matrix and list. Indeed,

data.frame is a list, with additional matrix-like attributes and it is always named (column names). is.list(data.frame) will return true. Therefore, [ (one indices vector), [[, $ behave exactly the same as a list. Thus, data.frame[1], data.frame[[1]] returns the first column, but one is a data.frame of one column, and the other is vector of first column. > Note: Indeed, there is a difference between a data.frame and a list. data.frame is NOT a vector (similar to matrix) while list is. Try is.vector(data.frame) and is.vector(list).
data.frame[i, j] behaves identical to matrix[i,j], except when i is vector of integer of length 1, while j is vector of length > 1, e.g, i = 2 where matrix[2, 1:3] returns a vector but data.frame[2, 1:3] gives a data.frame. (Think about why this is reasonable? Hint, data.frame is a list)

Use `[`, `[[` or `$` as function to apply-functions

Example, get the first entry or first 2 entries of each element in a list. Use them, surround them by back-ticks `. Again, bear in mind the difference between these operators.

> ly = list(AA = c('a' = 1, 'b' = 2), BB = c(3, 4, 'a' = 5, 6))
> ly
$AA
a b 
1 2 
$BB
    a   
3 4 5 6 

# The name of 1st element is 'AA.a' as `[` retains names
> sapply(ly, `[`, 1)
AA.a   BB 
   1    3 
> sapply(ly, `[[`,1)
AA BB 
 1  3 
> sapply(ly, `[`, i = 1:2)
  AA BB
a  1  3
b  2  4
> sapply(ly, `[[`, 'a')
AA BB 
 1  5

Define `[`, `[[` or `$` for your own object

> lz <- structure(list(
        AA = c('a' = 1, 'b' = 2), 
        BB = c(3, 4, 'a' = 5, 6)), 
    class = 'myclass')
> lz[1] # use `[` method for list
$AA
a b 
1 2 

# re-define `[` to return the i-th entry of each element of a list
> `[.myclass` <- function(x, i) sapply(x, `[[`, i)
> lz[1]
AA BB 
 1  3 
> lz['a']
AA BB 
 1  5

R variable assignment and copy

Sat, 15 Aug 2015 00:00:00 +0000

Assignment: copy-on-modify

When you do y = x, y only binds to the memory of data of x. R does not actually allocate new memory and copy x to y. But if you then do x[1] <- 1, R detects the modification and allocate memory for x by copy so that y is not changed. R is said to copy-on-modify. There is a package for checking memory address and number of bindings.

> library(pryr)
> x <- 1:10
> c(address(x), refs(x))
> [1] "0x103100060" "1"

> y <- x
> c(address(y), refs(y))
[1] "0x103100060" "2"

> x <- 1:10
> y <- x
> c(address(x), address(y))
[1] "0x6336498" "0x6336498"

> x[5] <- 6L
> c(address(x), address(y))
[1] "0x63b5318" "0x6336498"

You can see that y, x have the same address, but it has 2 names (x,y) binding to the same memory chunk in the first example (1st example), and addresses differ when one changed (2nd example).

Argument passing: pass-by-promise

For argument passing in R, it is pass-by-promise (un-evaluated expression), and it is evaluated to the actual data only when it is needed (R is lazy and smart!). Sometimes, you want an un-evaluated expression and you call substitute(x) inside the function, so that you can manipulated the expression. Check match.call as an example.

> # Make a BIG matrix
> x = matrix(runif(10000*10000), 10000);
> y = x; 
> # very fast! because no memory allocation happens.
> y[1,1] = -1;  
> # very slow, because y is re-allocated and copy happens.

> address(x)
[1] "0x7f94508e7010"

> ff = function(z)  address(z)
> ff(x)  # very fast and you know it is not copied. 
[1] "0x3ab6798"

Why not the same? I guess that is the address of the promise.

> fff = function(z)  {z2 = z; address(z2)} 
> fff(x)  # evaluate promise z to z2, but not copied
[1] "0x7f94508e7010"

Awesome, it is the same as x now! fff "evaluates" promise of z to z2. Indeed, what R does is just binding the memory to z2.

Example Post Formatting

Mon, 15 Dec 2014 00:00:00 +0000

Curabitur blandit tempus porttitor. Nullam quis risus eget urna mollis ornare vel eu leo. Vestibulum id ligula porta felis euismod semper. Donec sed odio dui. Aenean lacinia bibendum nulla sed consectetur.

Heading 1

Heading 2

Heading 3

Heading 4

Heading 5

Heading 6

Aenean lacinia bibendum nulla sed consectetur. Morbi leo risus, porta ac consectetur ac, vestibulum at eros. Cras mattis consectetur purus sit amet fermentum. Nulla vitae elit libero, a pharetra augue. Curabitur blandit tempus porttitor. Donec sed odio dui. Cras mattis consectetur purus sit amet fermentum.

Nullam quis risus eget urna mollis ornare vel eu leo. Cras mattis consectetur purus sit amet fermentum. Duis mollis, est non commodo luctus, nisi erat porttitor ligula, eget lacinia odio sem nec elit. Vivamus sagittis lacus vel augue laoreet rutrum faucibus dolor auctor.

Unordered List

List Item
Longer List Item
List Item

Ordered List

List Item
Longer List Item
List Item

Definition List

Coffee: Black hot drink
Milk: White cold drink

Donec id elit non mi porta gravida at eget metus. Cum sociis natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Maecenas faucibus mollis interdum. Donec sed odio dui. Cras justo odio, dapibus ac facilisis in, egestas eget quam.

Cras justo odio, dapibus ac facilisis in, egestas eget quam. Curabitur blandit tempus porttitor. Cum sociis natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Donec id elit non mi porta gravida at eget metus. Aenean eu leo quam. Pellentesque ornare sem lacinia quam venenatis vestibulum. Sed posuere consectetur est at lobortis. Vivamus sagittis lacus vel augue laoreet rutrum faucibus dolor auctor.

Maecenas faucibus mollis interdum. Maecenas faucibus mollis interdum. Duis mollis, est non commodo luctus, nisi erat porttitor ligula, eget lacinia odio sem nec elit. Etiam porta sem malesuada magna mollis euismod. Vestibulum id ligula porta felis euismod semper. Cras mattis consectetur purus sit amet fermentum.

Sed posuere consectetur est at lobortis. Fusce dapibus, tellus ac cursus commodo, tortor mauris condimentum nibh, ut fermentum massa justo sit amet risus. Fusce dapibus, tellus ac cursus commodo, tortor mauris condimentum nibh, ut fermentum massa justo sit amet risus. Aenean eu leo quam. Pellentesque ornare sem lacinia quam venenatis vestibulum.

Curabitur blandit tempus porttitor. Lorem ipsum dolor sit amet, consectetur adipiscing elit. Vivamus sagittis lacus vel augue laoreet rutrum faucibus dolor auctor. Curabitur blandit tempus porttitor. Nullam quis risus eget urna mollis ornare vel eu leo. Maecenas faucibus mollis interdum. Nullam id dolor id nibh ultricies vehicula ut id elit.

Featured Image

Sun, 14 Dec 2014 00:00:00 +0000

Lorem ipsum thor smash liege-bastogne-liege landbouwkrediet ombregt krabbe, rouleur derby is for lovers bonk giro gilbert bidon. Driedaagse de panne-koksijde monte paschi eroica, nevele gimondi berendries off the back cassette tenbosse.

Bahamontes lanterne rouge normandie belgium. Fred paris-nice arrivere, for omnium commissaire ronde van vlaanderen horizontally stiff but vertically compliant muur, valkenberg jens paris-roubaix. Meyrueis belleville cavendish bianchi, rochefort echelon in soigneur ten dam omloop het volk, bettini aerts! Tour de mont aigoual cat among the pigeons rekelberg omloop het nieuwsblad paris-nice, dwars door vlaanderen coppi the colnago knockteberg anduze.

Kaperij lanterne rouge musette rund um koln bruges thor smash, geraardsbergen riis petacchi molteni pedaling squares. Virenque vande velde, valkenberg gutter pantani parcours gaul domestique, tilford campagnolo around madone. Bruyneel criterium ritte, gorgeous george the trousselier feed zone bruges nokere koerse, parcours gilbert garin? Anquetil valkenberg bettini cat among the pigeons.

Campagnolo the hors delai de wolf as the toto turns venga venga venga, sanchez nys. Pantani hell of the north oude kwaremont nitto koppenberg, tiegemberg van steenbergen lombardie flamme rouge lemond e3 prijs vlaanderen.

Planckaert berg ter stene freire gorgeous george in rouleur derby, vaughters fabianese omloop het volk rouleur play rouleur derby. Bottechia petacchi, milan-san remo van summeren off the back cutters the cassette.

Nyvelocity pyrenees vande velde merckx. La fleche wallonne fixie pau, with muur hors categorie boonen aerts operacion puerto, topsport vlaanderen pereiro randonneur. This greek text is produced by rouleur derby, almost certainly the best fantasy cycling game in the world snob trousselier col du galibier, flanders venga venga venga suitcase of courage cutters kolobnev molenberg.

Post With A Code Snippet

Sat, 13 Dec 2014 00:00:00 +0000

You'll find this post in your `_posts` directory - edit this post and re-build (or run with the `-w` switch) to see your changes! To add new posts, simply add a file in the `_posts` directory that follows the convention: YYYY-MM-DD-name-of-post.ext.

Jekyll also offers powerful support for code snippets:

def print_hi(name)
  puts "Hi, #{name}"
end
print_hi('Tom')
#=> prints 'Hi, Tom' to STDOUT.

Check out the Jekyll docs for more info on how to get the most out of Jekyll. File all bugs/feature requests at Jekyll's GitHub repo.

Figure With A Caption

Fri, 12 Dec 2014 00:00:00 +0000

Fig1. - This is an example figcaption

<figure>
    <img src="/assets/img/touring.jpg" alt=""> 
    <figcaption>Fig1. - This is an example figcaption</figcaption>
</figure>

Image in Post

Thu, 11 Dec 2014 00:00:00 +0000

Test Post

Wed, 10 Dec 2014 00:00:00 +0000